正如我在上一篇文章中所说的，这里有一些c#代码，可以为任何大小的矩阵实现O(1)矩阵旋转。为了简洁性和可读性，没有错误检查或范围检查。代码:

static void Main (string [] args)
{
  int [,]
    //  create an arbitrary matrix
    m = {{0, 1}, {2, 3}, {4, 5}};

  Matrix
    //  create wrappers for the data
    m1 = new Matrix (m),
    m2 = new Matrix (m),
    m3 = new Matrix (m);

  //  rotate the matricies in various ways - all are O(1)
  m1.RotateClockwise90 ();
  m2.Rotate180 ();
  m3.RotateAnitclockwise90 ();

  //  output the result of transforms
  System.Diagnostics.Trace.WriteLine (m1.ToString ());
  System.Diagnostics.Trace.WriteLine (m2.ToString ());
  System.Diagnostics.Trace.WriteLine (m3.ToString ());
}

class Matrix
{
  enum Rotation
  {
    None,
    Clockwise90,
    Clockwise180,
    Clockwise270
  }

  public Matrix (int [,] matrix)
  {
    m_matrix = matrix;
    m_rotation = Rotation.None;
  }

  //  the transformation routines
  public void RotateClockwise90 ()
  {
    m_rotation = (Rotation) (((int) m_rotation + 1) & 3);
  }

  public void Rotate180 ()
  {
    m_rotation = (Rotation) (((int) m_rotation + 2) & 3);
  }

  public void RotateAnitclockwise90 ()
  {
    m_rotation = (Rotation) (((int) m_rotation + 3) & 3);
  }

  //  accessor property to make class look like a two dimensional array
  public int this [int row, int column]
  {
    get
    {
      int
        value = 0;

      switch (m_rotation)
      {
      case Rotation.None:
        value = m_matrix [row, column];
        break;

      case Rotation.Clockwise90:
        value = m_matrix [m_matrix.GetUpperBound (0) - column, row];
        break;

      case Rotation.Clockwise180:
        value = m_matrix [m_matrix.GetUpperBound (0) - row, m_matrix.GetUpperBound (1) - column];
        break;

      case Rotation.Clockwise270:
        value = m_matrix [column, m_matrix.GetUpperBound (1) - row];
        break;
      }

      return value;
    }

    set
    {
      switch (m_rotation)
      {
      case Rotation.None:
        m_matrix [row, column] = value;
        break;

      case Rotation.Clockwise90:
        m_matrix [m_matrix.GetUpperBound (0) - column, row] = value;
        break;

      case Rotation.Clockwise180:
        m_matrix [m_matrix.GetUpperBound (0) - row, m_matrix.GetUpperBound (1) - column] = value;
        break;

      case Rotation.Clockwise270:
        m_matrix [column, m_matrix.GetUpperBound (1) - row] = value;
        break;
      }
    }
  }

  //  creates a string with the matrix values
  public override string ToString ()
  {
    int
      num_rows = 0,
      num_columns = 0;

    switch (m_rotation)
    {
    case Rotation.None:
    case Rotation.Clockwise180:
      num_rows = m_matrix.GetUpperBound (0);
      num_columns = m_matrix.GetUpperBound (1);
      break;

    case Rotation.Clockwise90:
    case Rotation.Clockwise270:
      num_rows = m_matrix.GetUpperBound (1);
      num_columns = m_matrix.GetUpperBound (0);
      break;
    }

    StringBuilder
      output = new StringBuilder ();

    output.Append ("{");

    for (int row = 0 ; row <= num_rows ; ++row)
    {
      if (row != 0)
      {
        output.Append (", ");
      }

      output.Append ("{");

      for (int column = 0 ; column <= num_columns ; ++column)
      {
        if (column != 0)
        {
          output.Append (", ");
        }

        output.Append (this [row, column].ToString ());
      }

      output.Append ("}");
    }

    output.Append ("}");

    return output.ToString ();
  }

  int [,]
    //  the original matrix
    m_matrix;

  Rotation
    //  the current view of the matrix
    m_rotation;
}

好的，我把手举起来，当旋转时，它实际上不会对原始数组做任何修改。但是，在面向对象系统中，只要对象看起来像是被旋转到类的客户端，这就无关紧要了。目前，Matrix类使用对原始数组数据的引用，因此改变m1的任何值也将改变m2和m3。对构造函数稍加更改，创建一个新数组并将值复制到该数组中，就可以将其整理出来。

矩阵转置和旋转(+/-90，+/-180)的C代码

支持方阵和非方阵，具有原位和复制功能 支持2D数组和带有逻辑行/cols的1D指针 单元测试;有关使用示例，请参阅测试 测试gcc -std=c90 -Wall -pedantic, MSVC17

`

#include <stdlib.h>
#include <memory.h>
#include <assert.h>

/* 
    Matrix transpose & rotate (+/-90, +/-180)
        Supports both 2D arrays and 1D pointers with logical rows/cols
        Supports square and non-square matrices, has in-place and copy features
        See tests for examples of usage
    tested gcc -std=c90 -Wall -pedantic, MSVC17
*/

typedef int matrix_data_t;  /* matrix data type */

void transpose(const matrix_data_t* src, matrix_data_t* dst, int rows, int cols);
void transpose_inplace(matrix_data_t* data, int n );
void rotate(int direction, const matrix_data_t* src, matrix_data_t* dst, int rows, int cols);
void rotate_inplace(int direction, matrix_data_t* data, int n);
void reverse_rows(matrix_data_t* data, int rows, int cols);
void reverse_cols(matrix_data_t* data, int rows, int cols);

/* test/compare fn */
int test_cmp(const matrix_data_t* lhs, const matrix_data_t* rhs, int rows, int cols );

/* TESTS/USAGE */
void transpose_test() {

    matrix_data_t sq3x3[9] = { 0,1,2,3,4,5,6,7,8 };/* 3x3 square, odd length side */
    matrix_data_t sq3x3_cpy[9];
    matrix_data_t sq3x3_2D[3][3] = { { 0,1,2 },{ 3,4,5 },{ 6,7,8 } };/* 2D 3x3 square */
    matrix_data_t sq3x3_2D_copy[3][3];

    /* expected test values */
    const matrix_data_t sq3x3_orig[9] = { 0,1,2,3,4,5,6,7,8 };
    const matrix_data_t sq3x3_transposed[9] = { 0,3,6,1,4,7,2,5,8};

    matrix_data_t sq4x4[16]= { 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15 };/* 4x4 square, even length*/
    const matrix_data_t sq4x4_orig[16] = { 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15 };
    const matrix_data_t sq4x4_transposed[16] = { 0,4,8,12,1,5,9,13,2,6,10,14,3,7,11,15 };

    /* 2x3 rectangle */
    const matrix_data_t r2x3_orig[6] = { 0,1,2,3,4,5 };
    const matrix_data_t r2x3_transposed[6] = { 0,3,1,4,2,5 };
    matrix_data_t r2x3_copy[6];

    matrix_data_t r2x3_2D[2][3] = { {0,1,2},{3,4,5} };  /* 2x3 2D rectangle */
    matrix_data_t r2x3_2D_t[3][2];

    /* matrix_data_t r3x2[6] = { 0,1,2,3,4,5 }; */
    matrix_data_t r3x2_copy[6];
    /* 3x2 rectangle */
    const matrix_data_t r3x2_orig[6] = { 0,1,2,3,4,5 };
    const matrix_data_t r3x2_transposed[6] = { 0,2,4,1,3,5 };

    matrix_data_t r6x1[6] = { 0,1,2,3,4,5 };    /* 6x1 */
    matrix_data_t r6x1_copy[6];

    matrix_data_t r1x1[1] = { 0 };  /*1x1*/
    matrix_data_t r1x1_copy[1];

    /* 3x3 tests, 2D array tests */
    transpose_inplace(sq3x3, 3);    /* transpose in place */
    assert(!test_cmp(sq3x3, sq3x3_transposed, 3, 3));
    transpose_inplace(sq3x3, 3);    /* transpose again */
    assert(!test_cmp(sq3x3, sq3x3_orig, 3, 3));

    transpose(sq3x3, sq3x3_cpy, 3, 3);  /* transpose copy 3x3*/
    assert(!test_cmp(sq3x3_cpy, sq3x3_transposed, 3, 3));

    transpose((matrix_data_t*)sq3x3_2D, (matrix_data_t*)sq3x3_2D_copy, 3, 3);   /* 2D array transpose/copy */
    assert(!test_cmp((matrix_data_t*)sq3x3_2D_copy, sq3x3_transposed, 3, 3));
    transpose_inplace((matrix_data_t*)sq3x3_2D_copy, 3);    /* 2D array transpose in place */
    assert(!test_cmp((matrix_data_t*)sq3x3_2D_copy, sq3x3_orig, 3, 3));

    /* 4x4 tests */
    transpose_inplace(sq4x4, 4);    /* transpose in place */
    assert(!test_cmp(sq4x4, sq4x4_transposed, 4,4));
    transpose_inplace(sq4x4, 4);    /* transpose again */
    assert(!test_cmp(sq4x4, sq4x4_orig, 3, 3));

    /* 2x3,3x2 tests */
    transpose(r2x3_orig, r2x3_copy, 2, 3);
    assert(!test_cmp(r2x3_copy, r2x3_transposed, 3, 2));

    transpose(r3x2_orig, r3x2_copy, 3, 2);
    assert(!test_cmp(r3x2_copy, r3x2_transposed, 2,3));

    /* 2D array */
    transpose((matrix_data_t*)r2x3_2D, (matrix_data_t*)r2x3_2D_t, 2, 3);
    assert(!test_cmp((matrix_data_t*)r2x3_2D_t, r2x3_transposed, 3,2));

    /* Nx1 test, 1x1 test */
    transpose(r6x1, r6x1_copy, 6, 1);
    assert(!test_cmp(r6x1_copy, r6x1, 1, 6));

    transpose(r1x1, r1x1_copy, 1, 1);
    assert(!test_cmp(r1x1_copy, r1x1, 1, 1));

}

void rotate_test() {

    /* 3x3 square */
    const matrix_data_t sq3x3[9] = { 0,1,2,3,4,5,6,7,8 };
    const matrix_data_t sq3x3_r90[9] = { 6,3,0,7,4,1,8,5,2 };
    const matrix_data_t sq3x3_180[9] = { 8,7,6,5,4,3,2,1,0 };
    const matrix_data_t sq3x3_l90[9] = { 2,5,8,1,4,7,0,3,6 };
    matrix_data_t sq3x3_copy[9];

    /* 3x3 square, 2D */
    matrix_data_t sq3x3_2D[3][3] = { { 0,1,2 },{ 3,4,5 },{ 6,7,8 } };

    /* 4x4, 2D */
    matrix_data_t sq4x4[4][4] = { { 0,1,2,3 },{ 4,5,6,7 },{ 8,9,10,11 },{ 12,13,14,15 } };
    matrix_data_t sq4x4_copy[4][4];
    const matrix_data_t sq4x4_r90[16] = { 12,8,4,0,13,9,5,1,14,10,6,2,15,11,7,3 };
    const matrix_data_t sq4x4_l90[16] = { 3,7,11,15,2,6,10,14,1,5,9,13,0,4,8,12 };
    const matrix_data_t sq4x4_180[16] = { 15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0 };

    matrix_data_t r6[6] = { 0,1,2,3,4,5 };  /* rectangle with area of 6 (1x6,2x3,3x2, or 6x1) */
    matrix_data_t r6_copy[6];
    const matrix_data_t r1x6_r90[6] = { 0,1,2,3,4,5 };
    const matrix_data_t r1x6_l90[6] = { 5,4,3,2,1,0 };
    const matrix_data_t r1x6_180[6] = { 5,4,3,2,1,0 };

    const matrix_data_t r2x3_r90[6] = { 3,0,4,1,5,2 };
    const matrix_data_t r2x3_l90[6] = { 2,5,1,4,0,3 };
    const matrix_data_t r2x3_180[6] = { 5,4,3,2,1,0 };

    const matrix_data_t r3x2_r90[6] = { 4,2,0,5,3,1 };
    const matrix_data_t r3x2_l90[6] = { 1,3,5,0,2,4 };
    const matrix_data_t r3x2_180[6] = { 5,4,3,2,1,0 };

    const matrix_data_t r6x1_r90[6] = { 5,4,3,2,1,0 };
    const matrix_data_t r6x1_l90[6] = { 0,1,2,3,4,5 };
    const matrix_data_t r6x1_180[6] = { 5,4,3,2,1,0 };

    /* sq3x3 tests */
    rotate(90, sq3x3, sq3x3_copy, 3, 3);    /* +90 */
    assert(!test_cmp(sq3x3_copy, sq3x3_r90, 3, 3));
    rotate(-90, sq3x3, sq3x3_copy, 3, 3);   /* -90 */
    assert(!test_cmp(sq3x3_copy, sq3x3_l90, 3, 3));
    rotate(180, sq3x3, sq3x3_copy, 3, 3);   /* 180 */
    assert(!test_cmp(sq3x3_copy, sq3x3_180, 3, 3));
    /* sq3x3 in-place rotations */
    memcpy( sq3x3_copy, sq3x3, 3 * 3 * sizeof(matrix_data_t));
    rotate_inplace(90, sq3x3_copy, 3);
    assert(!test_cmp(sq3x3_copy, sq3x3_r90, 3, 3));
    rotate_inplace(-90, sq3x3_copy, 3);
    assert(!test_cmp(sq3x3_copy, sq3x3, 3, 3)); /* back to 0 orientation */
    rotate_inplace(180, sq3x3_copy, 3);
    assert(!test_cmp(sq3x3_copy, sq3x3_180, 3, 3));
    rotate_inplace(-180, sq3x3_copy, 3);
    assert(!test_cmp(sq3x3_copy, sq3x3, 3, 3));
    rotate_inplace(180, (matrix_data_t*)sq3x3_2D, 3);/* 2D test */
    assert(!test_cmp((matrix_data_t*)sq3x3_2D, sq3x3_180, 3, 3));

    /* sq4x4 */
    rotate(90, (matrix_data_t*)sq4x4, (matrix_data_t*)sq4x4_copy, 4, 4);
    assert(!test_cmp((matrix_data_t*)sq4x4_copy, sq4x4_r90, 4, 4));
    rotate(-90, (matrix_data_t*)sq4x4, (matrix_data_t*)sq4x4_copy, 4, 4);
    assert(!test_cmp((matrix_data_t*)sq4x4_copy, sq4x4_l90, 4, 4));
    rotate(180, (matrix_data_t*)sq4x4, (matrix_data_t*)sq4x4_copy, 4, 4);
    assert(!test_cmp((matrix_data_t*)sq4x4_copy, sq4x4_180, 4, 4));

    /* r6 as 1x6 */
    rotate(90, r6, r6_copy, 1, 6);
    assert(!test_cmp(r6_copy, r1x6_r90, 1, 6));
    rotate(-90, r6, r6_copy, 1, 6);
    assert(!test_cmp(r6_copy, r1x6_l90, 1, 6));
    rotate(180, r6, r6_copy, 1, 6);
    assert(!test_cmp(r6_copy, r1x6_180, 1, 6));

    /* r6 as 2x3 */
    rotate(90, r6, r6_copy, 2, 3);
    assert(!test_cmp(r6_copy, r2x3_r90, 2, 3));
    rotate(-90, r6, r6_copy, 2, 3);
    assert(!test_cmp(r6_copy, r2x3_l90, 2, 3));
    rotate(180, r6, r6_copy, 2, 3);
    assert(!test_cmp(r6_copy, r2x3_180, 2, 3));

    /* r6 as 3x2 */
    rotate(90, r6, r6_copy, 3, 2);
    assert(!test_cmp(r6_copy, r3x2_r90, 3, 2));
    rotate(-90, r6, r6_copy, 3, 2);
    assert(!test_cmp(r6_copy, r3x2_l90, 3, 2));
    rotate(180, r6, r6_copy, 3, 2);
    assert(!test_cmp(r6_copy, r3x2_180, 3, 2));

    /* r6 as 6x1 */
    rotate(90, r6, r6_copy, 6, 1);
    assert(!test_cmp(r6_copy, r6x1_r90, 6, 1));
    rotate(-90, r6, r6_copy, 6, 1);
    assert(!test_cmp(r6_copy, r6x1_l90, 6, 1));
    rotate(180, r6, r6_copy, 6, 1);
    assert(!test_cmp(r6_copy, r6x1_180, 6, 1));
}

/* test comparison fn, return 0 on match else non zero */
int test_cmp(const matrix_data_t* lhs, const matrix_data_t* rhs, int rows, int cols) {

    int r, c;

    for (r = 0; r < rows; ++r) {
        for (c = 0; c < cols; ++c) {
            if ((lhs + r * cols)[c] != (rhs + r * cols)[c])
                return -1;
        }
    }
    return 0;
}

/*
Reverse values in place of each row in 2D matrix data[rows][cols] or in 1D pointer with logical rows/cols
[A B C] ->  [C B A]
[D E F]     [F E D]
*/
void reverse_rows(matrix_data_t* data, int rows, int cols) {

    int r, c;
    matrix_data_t temp;
    matrix_data_t* pRow = NULL;

    for (r = 0; r < rows; ++r) {
        pRow = (data + r * cols);
        for (c = 0; c < (int)(cols / 2); ++c) { /* explicit truncate */
            temp = pRow[c];
            pRow[c] = pRow[cols - 1 - c];
            pRow[cols - 1 - c] = temp;
        }
    }
}

/*
Reverse values in place of each column in 2D matrix data[rows][cols] or in 1D pointer with logical rows/cols
[A B C] ->  [D E F]
[D E F]     [A B C]
*/
void reverse_cols(matrix_data_t* data, int rows, int cols) {

    int r, c;
    matrix_data_t temp;
    matrix_data_t* pRowA = NULL;
    matrix_data_t* pRowB = NULL;

    for (c = 0; c < cols; ++c) {
        for (r = 0; r < (int)(rows / 2); ++r) { /* explicit truncate */
            pRowA = data + r * cols;
            pRowB = data + cols * (rows - 1 - r);
            temp = pRowA[c];
            pRowA[c] = pRowB[c];
            pRowB[c] = temp;
        }
    }
}

/* Transpose NxM matrix to MxN matrix in O(n) time */
void transpose(const matrix_data_t* src, matrix_data_t* dst, int N, int M) {

    int i;
    for (i = 0; i<N*M; ++i) dst[(i%M)*N + (i / M)] = src[i];    /* one-liner version */

    /*
    expanded version of one-liner:  calculate XY based on array index, then convert that to YX array index
    int i,j,x,y;
    for (i = 0; i < N*M; ++i) {
    x = i % M;
    y = (int)(i / M);
    j = x * N + y;
    dst[j] = src[i];
    }
    */

    /*
    nested for loop version
    using ptr arithmetic to get proper row/column
    this is really just dst[col][row]=src[row][col]

    int r, c;

    for (r = 0; r < rows; ++r) {
        for (c = 0; c < cols; ++c) {
            (dst + c * rows)[r] = (src + r * cols)[c];
        }
    }
    */
}

/*
Transpose NxN matrix in place
*/
void transpose_inplace(matrix_data_t* data, int N ) {

    int r, c;
    matrix_data_t temp;

    for (r = 0; r < N; ++r) {
        for (c = r; c < N; ++c) { /*start at column=row*/
                                    /* using ptr arithmetic to get proper row/column */
                                    /* this is really just
                                    temp=dst[col][row];
                                    dst[col][row]=src[row][col];
                                    src[row][col]=temp;
                                    */
            temp = (data + c * N)[r];
            (data + c * N)[r] = (data + r * N)[c];
            (data + r * N)[c] = temp;
        }
    }
}

/*
Rotate 1D or 2D src matrix to dst matrix in a direction (90,180,-90)
Precondition:  src and dst are 2d matrices with dimensions src[rows][cols] and dst[cols][rows] or 1D pointers with logical rows/cols
*/
void rotate(int direction, const matrix_data_t* src, matrix_data_t* dst, int rows, int cols) {

    switch (direction) {
    case -90:
        transpose(src, dst, rows, cols);
        reverse_cols(dst, cols, rows);
        break;
    case 90:
        transpose(src, dst, rows, cols);
        reverse_rows(dst, cols, rows);
        break;
    case 180:
    case -180:
        /* bit copy to dst, use in-place reversals */
        memcpy(dst, src, rows*cols*sizeof(matrix_data_t));
        reverse_cols(dst, cols, rows);
        reverse_rows(dst, cols, rows);
        break;
    }
}

/*
Rotate array in a direction.
Array must be NxN 2D or 1D array with logical rows/cols
Direction can be (90,180,-90,-180)
*/
void rotate_inplace( int direction, matrix_data_t* data, int n) {

    switch (direction) {
    case -90:
        transpose_inplace(data, n);
        reverse_cols(data, n, n);
        break;
    case 90:
        transpose_inplace(data, n);
        reverse_rows(data, n, n);
        break;
    case 180:
    case -180:
        reverse_cols(data, n, n);
        reverse_rows(data, n, n);
        break;
    }
}

`

基于大量的其他答案，我用c#想出了这个:

/// <param name="rotation">The number of rotations (if negative, the <see cref="Matrix{TValue}"/> is rotated counterclockwise; 
/// otherwise, it's rotated clockwise). A single (positive) rotation is equivalent to 90° or -270°; a single (negative) rotation is 
/// equivalent to -90° or 270°. Matrices may be rotated by 90°, 180°, or 270° only (or multiples thereof).</param>
/// <returns></returns>
public Matrix<TValue> Rotate(int rotation)
{
    var result = default(Matrix<TValue>);

    //This normalizes the requested rotation (for instance, if 10 is specified, the rotation is actually just +-2 or +-180°, but all 
    //correspond to the same rotation).
    var d = rotation.ToDouble() / 4d;
    d = d - (int)d;

    var degree = (d - 1d) * 4d;

    //This gets the type of rotation to make; there are a total of four unique rotations possible (0°, 90°, 180°, and 270°).
    //Each correspond to 0, 1, 2, and 3, respectively (or 0, -1, -2, and -3, if in the other direction). Since
    //1 is equivalent to -3 and so forth, we combine both cases into one. 
    switch (degree)
    {
        case -3:
        case +1:
            degree = 3;
            break;
        case -2:
        case +2:
            degree = 2;
            break;
        case -1:
        case +3:
            degree = 1;
            break;
        case -4:
        case  0:
        case +4:
            degree = 0;
            break;
    }
    switch (degree)
    {
        //The rotation is 0, +-180°
        case 0:
        case 2:
            result = new TValue[Rows, Columns];
            break;
        //The rotation is +-90°
        case 1:
        case 3:
            result = new TValue[Columns, Rows];
            break;
    }

    for (uint i = 0; i < Columns; ++i)
    {
        for (uint j = 0; j < Rows; ++j)
        {
            switch (degree)
            {
                //If rotation is 0°
                case 0:
                    result._values[j][i] = _values[j][i];
                    break;
                //If rotation is -90°
                case 1:
                    //Transpose, then reverse each column OR reverse each row, then transpose
                    result._values[i][j] = _values[j][Columns - i - 1];
                    break;
                //If rotation is +-180°
                case 2:
                    //Reverse each column, then reverse each row
                    result._values[(Rows - 1) - j][(Columns - 1) - i] = _values[j][i];
                    break;
                //If rotation is +90°
                case 3:
                    //Transpose, then reverse each row
                    result._values[i][j] = _values[Rows - j - 1][i];
                    break;
            }
        }
    }
    return result;
}

其中_values对应于由Matrix<TValue>定义的私有二维数组(形式为[][])。result = new TValue[Columns, Rows]可能通过隐式操作符重载并将二维数组转换为Matrix<TValue>。 Columns和Rows两个属性是公共属性，用于获取当前实例的列数和行数:

public uint Columns 
    => (uint)_values[0].Length;

public uint Rows 
    => (uint)_values.Length;

当然，假设您更喜欢使用无符号下标;-)

所有这些都允许您指定它应该旋转多少次，以及它应该向左旋转(如果小于零)还是向右旋转(如果大于零)。您可以改进此方法，以检查实际角度的旋转，但如果值不是90的倍数，则可能会抛出异常。有了这些输入，你可以相应地改变方法:

public Matrix<TValue> Rotate(int rotation)
{
    var _rotation = (double)rotation / 90d;

    if (_rotation - Math.Floor(_rotation) > 0)
    {
        throw new NotSupportedException("A matrix may only be rotated by multiples of 90.").
    }

    rotation = (int)_rotation;
    ...
}

Since a degree is more accurately expressed by double than int, but a matrix can only rotate in multiples of 90, it is far more intuitive to make the argument correspond to something else that can be accurately represented by the data structure used. int is perfect because it can tell you how many times to rotate it up to a certain unit (90) as well as the direction. double may very well be able to tell you that also, but it also includes values that aren't supported by this operation (which is inherently counter-intuitive).

当前所有的解决方案都有O(n^2)开销作为临时空间(这不包括那些肮脏的OOP骗子!)这里有一个内存占用为O(1)的解决方案，将矩阵原地右转90度。该死的延展性，这玩意儿跑得很快!

#include <algorithm>
#include <cstddef>

// Rotates an NxN matrix of type T 90 degrees to the right.
template <typename T, size_t N>
void rotate_matrix(T (&matrix)[N][N])
{
    for(size_t i = 0; i < N; ++i)
        for(size_t j = 0; j <= (N-i); ++j)
            std::swap(matrix[i][j], matrix[j][i]);
}

免责声明:我实际上并没有测试这个。让我们玩打虫游戏吧!

可以做递归相当干净，这里是我的实现在golang!

在没有额外内存的情况下递归地旋转go golang中的NXN矩阵

func rot90(a [][]int) {
    n := len(a)
    if n == 1 {
        return
    }
    for i := 0; i < n; i++ {
        a[0][i], a[n-1-i][n-1] = a[n-1-i][n-1], a[0][i]
    }
    rot90(a[1:])
}

O(n²)时间和O(1)空间算法(没有任何变通方法和恶作剧的东西!)

旋转+90:

转置 反转每行

旋转-90:

方法一:

转置 反转每一列

方法二:

反转每行 转置

旋转180度:

方法一:旋转+90两次

方法2:反转每行，然后反转每列(转置)

旋转-180度:

方法一:旋转-90度2次

方法二:先反转每一列，再反转每一行

方法三:旋转+180，因为它们是相同的